PIMS-BIRS TeamUp: Statistical Methodology for Model Checking for Aggregated Relational Data (25frg505)

Organizers

Owen Ward (SFU)

Description

The Banff International Research Station will host the PIMS-BIRS TeamUp: "Statistical Methodology for Model Checking for Aggregated Relational Data" workshop at UBCO Kelowna from June 1 - 15, 2025.


Aggregated Relational Data (ARD) contains information about the number of connections an actor has to other actors with a given trait. The ARD are commonly collected by asking respondents questions of the form, “How many X’s do you know?”, where X represents some traits of interest. Researchers have used ARD to estimate the size of hidden populations like female sex workers, study sociological behaviors of groups such as the typical individual’s average number of social connections, and estimate properties of complete networks like how socioeconomic outcomes affect network features. ARD is often preferable to complete network data because researchers can collect ARD at only a fraction of the cost and effort of collecting a complete network. However, while substantial work has been done on developing new ARD models to further explore different properties of the data, very little discussion has focused on how well ARD models actually fit the data and how the privacy of the respondents can be preserved while still yielding valid inference? This project lays the foundation for this research direction by providing researchers with the first approach to perform a rigorous and principled analysis of statistical models for this data. This will have important consequences and may be used for tasks such as protecting the identity of survey respondents. Together, these developments allow both statisticians and practitioners to collect and analyze ARD in a safe and reliable way.


The Banff International Research Station for Mathematical Innovation and Discovery (BIRS) is a collaborative Canada-US-Mexico venture that provides an environment for creative interaction as well as the exchange of ideas, knowledge, and methods within the Mathematical Sciences, with related disciplines and with industry. BIRS is supported by Canada’s Natural Science and Engineering Research Council (NSERC), the U.S. National Science Foundation (NSF), and Alberta’s Advanced Education and Technology.